Monday, 18 April 2016

World closer to ending in tears: customers being paid to have a mortgage

Fire-fighting is dangerous
Negative interest rates have seemed a somewhat cute theoretical issue to many consumers. Negative interest rates? Perhaps just for the big end of town?

You may have grumbled that someone must be making out like a bandit as you continue to pay your mortgage.

No longer.

A funny / scary story in the WSJ [Negative Rates Around the World, Duxbury & Gauthier-Villars, April 14, 2016] relates the personal story of a Danish bank mortgage customer being paid by his bank to have a mortgage.

Here are some choice sections from the WSJ article:
AALBORG, Denmark— Hans Peter Christensen got some unusual news when he opened his most recent mortgage statement. His quarterly interest payment was negative 249 Danish kroner. 
Instead of paying interest on the loan he got a decade ago to buy a house in this northern Denmark city, his bank paid him the equivalent of $38 in interest for the quarter. As of Dec. 31, his mortgage rate, excluding fees, stood at negative 0.0562%.
...
“My parents said I should frame it, to prove to coming generations that this ever happened,” said Mr. Christensen, a 35-year-old financial consultant, about his bank statement."
And some of the consequences:
Prices of owner-occupied apartments have been rising. In Copenhagen, prices were 14.5% higher in the fourth quarter of 2015 than in the year-earlier period, compared with a 5.5% increase in 2014, according to the Association of Danish Mortgage Banks. 
In Stockholm, prices rose 17% in 2015, following a 10% increase in 2014, according to price-tracking company Svensk Maklarstatistik AB.
This isn't going to end pretty.

--Matt.

Thursday, 14 April 2016

Solarflare is suing Exablaze for patent infringement

Solarflare issued a press release announcing the court action against Exablaze for patent infringement. I can't help but not be not silent about this one, so let's meander through...

Criminal firms, such as Exablaze, deserve bad karma
Solarflare are the good guys in this battle. Exablaze is the quasi-criminal, nasty, patent-troll-like organisation that deserves your scorn.

I've read through but have not done a full analysis of the four patents subject to Solarflare's infringement allegations. Most of the patents have a priority date significantly before Exablaze was created. I know this as I instigated and managed the first FPGA NIC development at Zomojo, the firm that owns and span out Exablaze. So Solarflare have good priority at least.

The patents in question are linked here:
  1. US8116312-B2: Method and apparatus for multicast packet reception, Feb 8, 2006
    Claim of using filter to bypass kernel for delivery to user endpoint.
  2. US8645558-B2: Reception according to a data transfer protocol of data directed to any of a plurality of destination entities for data extraction, Jun 15, 2005
    Relates to usurping the OS by a user, or higher than OS level mechanism, for directing network data to an endpoint.
  3. US8817784-B2: Method and apparatus for multicast packet reception, Feb 8, 2006
    A continuation of US8116312-B2 regarding the establishment of packet filtering.
  4. US9258390-B2: Reducing network latency, Jul 29, 2011
    Using templates in a particular way to set up for transmitting protocols, filling those bits in, and subsequently transmitting the completed network message.
Exablaze's CTO Matt Chapman has commented on Solarflare tech in the past so his familiarity with it may indeed open Exablaze up to triple damages.

It is easy for me to pick sides as my I've expended considerable energy fighting against Exablaze / Zomojo. Greg Robinson, Matt Chapman, and their lawyers, such as Matt Critchley, deserve gaol time for the lying to the court, obstruction of justice, and other criminal acts. Matt Chapman lied to the court about a myriad of over 100 confidential claims. Chapman even tried to claim Sir Issac Newton's interpolation method was a valuable and confidential secret, proprietary only to his firm,
MR WOOD: I want to ask you a question about paragraph 98 of your MC1 document?CHAPMAN:---Yes.
HER HONOUR: Sorry, what? I missed that paragraph number. I apologise.MR WOOD: 98, if your Honour pleases.HER HONOUR: Thank you.
MR WOOD: Who is the Newton referred to in paragraph 98?
CHAPMAN---I don’t know how that’s relevant. That’s – it’s a well-known method for numerical approximate. It’s not Isaac Newton, no, I don’t think, but it may be. I don’t see how that’s relevant.
MR WOOD: I’m suggesting to you that it was Sir Isaac Newton’s method that was first published in 1685?
CHAPMAN: ---Okay.
Do you agree with that proposition?---That’s feasible. I don’t know.
It’s basic mathematics, isn’t it?---I would agree with that, yes.
Secret & confidential! Chapman and Robinson couldn't lie straight in bed.

In my action against them they had the benefit of being a friend of the judge and employing barristers who worked alongside the judge during the day. Terrible bias. It didn't help that the matter was considered so important that the eventual appeal was tacked onto a long sojourn by matey judges where one of the Full Bench jurists, Honourable (sic) Justice Gilmour from WA, turned up intoxicated in court (4th March 2015). There was never a chance.

Without a corrupt Australian court process, I expect Solarflare may fare better in the District Court of New Jersey. Also, being a jury trial, it may fairer. The full bench of the Federal Court of Australia did make a ruling that Greg Robinson was lying in their written appeal judgment. Exablaze has a history of lying to customers, staff, and courts. Perhaps Exablaze's lack of credility will help Solarflare in their just battle. A NJ court certainly should discount everything Exablaze, Chapman, and Robinson have to say. They are known to lie to courts.

Getting back to the tech, Solarflare make the #1 trading oriented network cards in the world in my mind. Simply put, if you want a high performance card with no fuss, use Solarflare. Having said that, I typically use Mellanox cards as the performance for me has been slightly better, with an improved cost profile, but with gritted teeth. From time to time, Mellanox have been a pain to get working properly, especially with in house apps. Solarflare, for me, is the “goto” card as its driver, OpenOnload(R), has been so much “cleaner” to work with. The Solarflare driver was very nice tech and I presume it still is.

It was interesting to read the Exablaze response to the court action. After Exablaze spent many years in court, lying to the court, perverting justice to misappropriate the intellectual property and patents of others, and cynically exploiting court processes, they had the temerity to include the following line in their PR response,
“While we respect any company's right to protect and defend its IP, we sincerely hope that Solarflare's actions in this instance are not a cynical attempt to exploit the legal process, rather than use innovation as a strategy to advance its competitive position.”
Yeah, right. What an ass-hat Exablaze's Greg Robinson truly is.

I hope karma has its day and Solarflare succeed in punishing the criminal firm Exablaze in court. I
Asshole firms, like Exablaze, sometimes get the karma they deserve
would also hope other firms, such as Exegy and Mellanox, who are likely to have intellectual property being violated by Exablaze, also take action. Indeed patent law says they need to take action as they otherwise surrender their rights. It would serve the community well to drive scum like Exablaze out of the industry.

I feel sympathy for Exablaze's customers and distributors who have been duped by Exablaze. Those clients may now be subject to having to return or give-up their Exablaze products. Such action may be hard to swallow but it may end up being a good thing, as there are certainly much better vendors to deal with.

Happy trading,

--Matt.

Saturday, 2 April 2016

A simple transport plan - make a few billion - save a few billion

I've long been a fan of high speed rail. No longer.

Let's meander through why autonomous-only roads may be better than rail, save billions of dollars, and make billions of dollars.

I've not much personal experience of high-speed rail, but I like it. I've only travelled on the impressive Maglev in Shanghai to airport train and a Shinkansen on the Osaka to Tokyo route. Only a slow 300 km/h in Shanghai (it will do over 400 km/h) and just a bit short of 300 km/h in Japan. The Shinkansen E5 series now will do around 320 km/h.

High speed railways are expensive to build and very expensive to maintain. In Japan thousands of workers are constantly doing track inspections and maintenance. Despite the high costs, the economic benefits of over 300M passengers per year in Japan piling onto bullet trains are obvious. That's a lot of time saved. It has been estimated that alone saves some ¥500 billion per year.
JR West machines from 2008
(click to enlarge, source)
Europe has some productive high speed rail. The US has long pined for it but has not been able to generate the will for even modest speed rail. Australia has long flirted with the idea of an eastern seaboard or Sydney-Melbourne link without any realistic momentum. It may have been lucky both countries have failed to launch as, just as with any technology, disruptive new ideas could change the game.

Want to make a few billion?

Perhaps a way to generate a few billion in profits, especially around geographically challenged cities such as San Franciso or Sydney, would be to have an interest in land that is difficult to have a sane commute to. Arrange the land and build high speed rail to grease the value proposition.

This is commonly referred to as "value capture infrastructure." For example, say you built a 500 km/h Maglev commuter to Willows CA from down-town San Francisco. That's about 225km (140 miles) by road or a bit under 200km as the crow flies. See the following map. Willows CA is inside the second biggest circle to the north of SF.
San Francisco with 50, 100, 200 & 400 km rings
(click to enlarge)
Imagine you bought enough land for 100,000 homes around Willows or a similarly distanced locale. Now build your MagLev. Say it has a top speed of 500km/h and you get a commute of under 30 minutes. That's pretty convenient. If your land is suddenly worth $100,000 more per block, or maybe $200,000 more, then that is a nice increase in value of $10B-20B. I'm not sure this is a great example, but you get the idea.

Earthquakes aside, it makes the economics of a 1000km/h Hyperloop to "nowhere in particular" pretty interesting, no?

Grow your city to 1M blocks of land and there is a lazy hundred billion in economic value to share between gov't, developers, and their customers.

A better plan

Elon Musk's Model 3 and other cars with autonomous capabilities, electric or not, provide a simpler solution. Imagine roads for only autonomous vehicles. No humans, so why not allow 200km/h travel?

Build a private road of 100 km in length. Own the property around the end of it. Profit.

Not enough autonomous capable cars? Have your autonomous car capsules ranked up and ready to roll just like a tiny train. Small packets of people cut down the latency of travel through quick start times.

A condition of travel would be autonomous only at the road's gateways. I'd certainly appreciate the freedom to work and travel without worrying about my safety or the inconvenience of discontinuous public transportation.
Tesla Model 3

It will take a while for cars to be able to legally self-drive on all public roads for good reason. Whilst autonomous cars will save lives and be a massive net gain, they will also cost the occasional life that a human may have avoided. In this Age of Perception, with its attendant disruption, higher level artificial reasoning will be slow to arrive. You would slow down when a beach ball had rolled across the road, even if it was well in front of you.  Your high level risk assessment would suggest  a child may be likely to follow. This is not the kind of logic a computer will easily carry out for the foreseeable future. Saving a thousand lives for the cost of one may be a good trade-off but an unnecessary loss is hard to swallow. This is an example of the kind of thinking that necessarily complicates and slows full autonomous transport adoption.

Governments could immensely improve the economic efficiency, safety, and liveability of their cities by allowing protected autonomous only lanes or roads. It is not just an acceleration of autonomous driving economic benefits. It is tangibly different. The main thing is with the microsecond attention span of the computer it will be safer to travel at very high speeds. You don't want human unpredictability interference in an autonomous environment. That is, even with autonomous driving, I don't want my car travelling at 200 or 300 km/h if someone may disrupt my life fatally by glancing at their text message on their phone. The only real and safe pathway to high speed driving is to take the human out of the picture.

Another benefit over high speed rail is that you will not need a few thousand workers inspecting short sections of road all the time. Maintenance costs would be much lower than high speed rail. We know how to build autobahns and other high speed roads. Autonomous cars can adapt to conditions and other vehicles through local sensing and communication.

This is similar, in some regards, to the Personal Rapid Transport idea that has been floating around for a long time. Well, since Donn Fitcher's work in 1953.

Autonomous commercial driving of trucks and buses will eventually displace the >11M trucks and buses on US roads too. Over 3M commercial drivers in the US will be redundant with more than 400M road hours of driving saved (2013 stats). That's quite some disruption with some sadness attached to the displaced until they find new, and hopefully better, work. Whilst road deaths in the US have dropped from over 50k to just over 30k, that many people dying on the road is just way too many. However, with not just >30k deaths, but more than 2.3M people being injured on US roads, there is a moral imperative to hasten autonomous driving.

Sydney

Here is a map for the difficult terrain around my home in Sydney, Australia. Sydney's geography is such that it is one of the most expensive cities in the world to own property. Many suburbs, run-of-the-mill suburbs, have property prices over $AUD 1 million. Due to this property stress, many people commute for two or three hours a day just to get to work. Not much quality of life in that.
Sydney with 50km and 100km rings from the CBD.
(click to enlarge)
If you send a private autonomous road, or lane-way, toward Newcastle, Nowra, or Bathhurst, I'd imagine many people would be attracted by a short 30 minute commute in a 200km/h car from Katoomba, Lake Macquarie, Woolombi, Kiama, or Bowral with the savings in the price of housing. It would be faster than the existing trip for many people, like me, who live the burbs.

Wake up at 7:45am in Moss Vale. A quick shower and change. Jump in the car at 8AM. Get half an hour of work in on the autonomous commute into the city and get your early e-mails done. The car drops you off at the office. Your car then goes off and finds a park. After work, summon the car to the office. It picks you up and you relax watching the Nightly News on the way home and you're home by 5:45PM to spend some quality time with the family without the mortgage stress.

Perhaps it may be a Google-taxi or a Musk-pod? Maybe it is simply a government run electric car-pod for the main highway. It all sounds a big win to me.


Win win

Autonomous only roads. Good for governments. Good for developers. Good for car manufacturers. Good for freight. Good for consumers. It's good economics. Please get on with it.

--Matt.